Pharmacological Ascorbate Radiosensitizes Pancreatic Cancer

Abstract

The toxicity of pharmacologic ascorbate is mediated by the generation of H2O2 via the oxidation of ascorbate. Because pancreatic cancer cells are sensitive to H2O2 generated by ascorbate, they would also be expected to become sensitized to agents that increase oxidative damage such as ionizing radiation. The current study demonstrates that pharmacologic ascorbate enhances the cytotoxic effects of ionizing radiation as seen by decreased cell viability and clonogenic survival in all pancreatic cancer cell lines examined, but not in nontumorigenic pancreatic ductal epithelial cells. Ascorbate radiosensitization was associated with an increase in oxidative stress-induced DNA damage, which was reversed by catalase. In mice with established heterotopic and orthotopic pancreatic tumor xenografts, pharmacologic ascorbate combined with ionizing radiation decreased tumor growth and increased survival, without damaging the gastrointestinal tract or increasing systemic changes in parameters indicative of oxidative stress. Our results demonstrate the potential clinical utility of pharmacologic ascorbate as a radiosensitizer in the treatment of pancreatic cancer.

Conflict of interest statement

There are no conflicts of interest

©2015 American Association for Cancer Research.

Figures

Figure 1. Ascorbate radiosensitizes pancreatic cancer cells

A. MIA PaCa-2 pancreatic cancer cells were irradiated (0 – 3 Gy) with and without ascorbate (0.25 mM) and clonogenic survival determined. The resulting dose modification factor at 10% iso-survival was 2.5 indicating enhanced radiosensitivity (Means ± SEM, n = 3). Pharmacological ascorbate (0.25 mM) produces a flux of H2O2 of 550 amol−1 cell−1 s−1 under our experimental conditions. This rate is considerably greater than the rate of cellular oxygen consumption by these cells, 57 amol−1 cell−1 s−1 (29). B. PANC-1 pancreatic cancer cells were irradiated (0–10 Gy) and treated with ascorbate (0.25 mM) in a similar fashion. Clonogenic survival yielded a DMF of 2.2 indicating enhanced radiosensitivity. (Means ± SEM, n = 3). C. AsPC-1 human pancreatic cancer cells were irradiated (0 – 10 Gy) with and without ascorbate (0.25 mM) and clonogenic survival determined. The resulting dose modification factor at 10% iso-survival was 1.25 indicating enhanced radiosensitivity. (Means ± SEM, n = 3). D. H6c7 immortalized human pancreatic ductal epithelial cells were irradiated (0–6 Gy) and treated with ascorbate (0.25 mM) to determine clonogenic survival. The clonogenic survival assay demonstrates no radiosensitization after IR with or without ascorbate. (Means ± SEM, n = 3). E. MIA PaCa-2 pancreatic cancer cells were irradiated (0–4 Gy) and treated with ascorbate (0.25 mM) to determine viability using the MTT assay and demonstrated radiosensitization. Data were normalized to drug or vehicle control. (Means ± SEM, n = 3). F. H6c7 immortalized human pancreatic ductal epithelial cells were irradiated (0–10 Gy) and treated with ascorbate (0.25 mM) to determine cell viability. The MTT assay demonstrates minimal changes in cell viability after IR with or without ascorbate. (Means ± SEM, n = 3). G. Patient derived pancreatic cancer cells (339) in 4% O2 were irradiated (0 – 6 Gy) with and without ascorbate (0.5 mM) and clonogenic survival determined. The resulting dose modification factor at 40% iso-survival was 1.4 indicating enhanced radiosensitivity. (Means ± SEM, n =3). H. Patient derived pancreatic cancer cells (403) in 4% O2 were irradiated (0 – 6 Gy) with and without ascorbate (0.5 mM) and clonogenic survival determined. The resulting dose modification factor at 40% iso-survival was 1.6 indicating enhanced radiosensitivity. (Means ± SEM, n =3).

Figure 2. Ascorbate radiosensitization is mediated by H2O2

A. MIA PaCa-2 cells were treated with 3-amino-1,2,4-triazole (3-AT) to determine intracellular H2O2. Cells treated with IR + ascorbate (20 mM) showed an increased concentration of intracellular H2O2 (n = 3, means ± SEM, P < 0.01). Pharmacological ascorbate at 20 mM produces H2O2 at a flux of 6200 amol−1 cell−1 s−1 in our experimental conditions. B. Ascorbate (2 mM) radiosensitization depletes GSH and increases the half-cell reduction potential (Ehc). Ascorbate at 2 mM produces H2O2 at a flux of 1400 amol−1 cell−1 s−1 in our experimental conditions. (Means ± SEM, n =3). C. Catalase partially rescues ascorbate (2 mM) radiosensitization. Western blot demonstrates increased catalase immunoreactive protein in cells treated with the AdCAT vector compared to AdEmpty treated cells. D. MIA PaCa-2 cells were transfected with AdEmpty (50 MOI) or AdCAT (50 MOI) for 48 h, then subjected to ionizing radiation (2 Gy), ascorbate (2 mM) or the combined treatment. Ascorbate radiosensitization was reversed with catalase overexpression. (n = 3, means ± SEM, P < 0.01).

Figure 3. Ascorbate radiosensitization induces DNA damage

A. In MIA PaCa-2 pancreatic cancer cells, both IR (2 Gy) and ascorbate (2 mM) induced the formation of γ-H2AX as determined by Western blot, while the combination treatment further enhanced γ-H2AX formation. B. In AsPC-1 pancreatic cancer cells, ascorbate (2 mM) radiosensitization increased γH2AX immunoreactive protein, which was reversed with catalase pretreatment. C. γ-H2AX immunohistochemistry demonstrated increased γH2AX formation with the combination treatment, which was reversed with catalase pretreatment. Cells were treated with ascorbate (1 mM) and IR (1 Gy) or catalase (100 U/mL). Ascorbate at 1 mM produces H2O2 at a flux of 690 amol−1 cell−1 s−1 under our experimental conditions. D. Quantification of γ-H2AX immunohistochemistry in MIA PaCa-2 cells treated with ascorbate (1 mM) and IR (1 Gy) demonstrating increased induction of γ-H2AX during ascorbate radiosensitization which was reversed with catalase pretreatment. (n = 3, means ± SEM, P < 0.01).

Figure 4. Pharmacological ascorbate radiosensitization in vivo

A. Linear mixed effects regression models were used to estimate and compare group-specific tumor growth curves. Tumor growth was significantly inhibited with the ascorbate + IR treatment compared to control animals or animals that received ascorbate alone. Control was saline (1 M NaCl i.p. daily, 22.7 μL g−1); IR (7.5 Gy on days 5 and 8 and 1 M NaCl i.p. daily); ascorbate (4 g kg−1 i.p. daily); or ascorbate + IR. (Means ± SEM, n = 9–11 animals/group). B. Kaplan-Meier survival plots demonstrating survival as a function of time. The log-rank test was used for pairwise treatment group comparisons of survival between treatment groups demonstrating significantly increased overall survival of animals receiving pharmacological ascorbate and IR. C. Pharmacological ascorbate alters the status of the GSH redox buffer of RBCs. Blood was collected from separate groups of mice after treatments and assayed for the intracellular concentration of GSH in the RBCs. Both IR and ascorbate alone decreased intracellular GSH compared to controls. The combination of ascorbate and IR did not further decrease GSH when compared to IR alone. D. Complete blood counts and differential in control mice and those treated with ascorbate, IR and IR + ascorbate. White blood cell counts were decreased with IR which was unchanged when ascorbate was added to the treatment regimen. E. Initially, purified bovine serum albumin was reacted with excess purified 4-HNE to create a positively labelled protein control which is seen in the upper panel. The negative control is purified bovine serum albumin without purified 4-HNE. Then to determine 4-HNE modified proteins in separate groups of mice, dot blots for 4-HNE modified proteins in cardiac muscle showed no changes in immunoreactive protein after any of the treatments compared to controls. F. Dot blots for 4-HNE modified proteins in heart, kidney and liver demonstrated no changes in immuno-reactive protein after any of the treatments compared to controls.

Figure 5. Pharmacologic ascorbate radiosensitization in an orthotopic model

A. Ultrasound guided pancreatic injections. 4 x 105 MIA PaCa-2 cells suspended in a 20 μL 1:1 mixture of PBS and Matrigel were injected into athymic nude mice who were sedated with Isoflurane. Under ultrasound guidance, injections were performed directly into the pancreas just medial and inferior to the spleen. The spleen and pancreas are labeled. The yellow arrow indicates the shadow of the needle during injection. B. Bioluminescence imaging microscopy. Four days after the initial injections, mice were imaged using the Xenogen IVIS 200 microscope 10 minutes after injection with 200 μL of 15 mg/mL luciferin. Initial exposure time was 1 minute. Displayed is a color scale of photons/second emitted superimposed over a photograph of the mouse as viewed in the prone position. The bioluminescence is localized to the left quadrant central quadrant as would be expected for a pancreatic tumor. C. Linear mixed effects regression models demonstrated significant inhibition in tumor growth with the ascorbate + IR treatment compared to control animals or animals that received IR alone. Control was saline (1 M NaCl i.p. daily, 22.7 μL g−1); IR (7.5 Gy on days 4 and 11 and 1 M NaCl i.p. daily); ascorbate (4 g kg−1 i.p. daily); or ascorbate + IR. (Means, n = 7 – 12 animals/group). *P < 0.05 control vs. Ascorbate + IR; #P < 0.05 Ascorbate + IR vs. IR alone.

Figure 6. Pharmacological ascorbate partially reverses IR-induced jejunal damage

A. Crypt cell assay from control mice. Sections were made of the jejunum and then viewed under light microscopy and expressed as surviving cells/circumference. B. Mice were treated with ascorbate 4 g/kg I.P. for 5 days and then sacrificed. C. Mice were exposed to 10 Gy of total abdominal radiation. After 48 h, each animal was sacrificed and sections were made of the jejunum as described. D. Mice treated with pharmacological ascorbate (4 g/kg for 5 days with 10 Gy total abdominal radiation on day 3 and then sacrificed on day 5. E. Mice exposed to 13 Gy of total abdominal radiation and sacrificed 48 h later. Note marked decrease in number of regenerating crypts/circumference of the sectioned jejunum. F. Mice treated with pharmacological ascorbate (4 g/kg for 5 days with 13 Gy total abdominal radiation on day 3 and then sacrificed on day 5. Note increase in regenerating crypts compared to panel E. G. Quantification of crypt cell assay. Means ± SEM, n = 8–10 samples/group. * P < 0,01 vs. Controls and ascorbate treated animals. #P < 0.05 vs. Control and ascorbate treated animals.

Figure 7. Pharmacological ascorbate radiosensitization in vivo

A. Linear mixed effects regression models were used to estimate and compare group-specific tumor growth curves. Tumor growth was significantly inhibited in mice with pancreatic tumor xenografts treated with IR/gemcitabine compared to controls and ascorbate alone. Tumor growth was further inhibited in mice that received the ascorbate + IR/gemcitabine treatment compared to control animals or animals that received ascorbate alone. (Means ± SEM, n = 11–12 mice/group). B. Kaplan-Meier survival plots demonstrating survival as a function of time. The log-rank test was used for pairwise treatment group comparisons of survival between treatment groups demonstrating significantly increased overall survival animals receiving pharmacological ascorbate and IR/gemcitabine treatment on day 43 after the initiation of treatment. C. Weight changes of mice during treatment periods in mice during treatments from day 1 through day 15.